The present relates to a method of regulating the rotational speed of multicylinder internal combustion engines having a crankshaft and fuel injection, in particular slow-running marine diesel engines, at least one sensor is used to detect the speed and direction of rotation of a pole wheel, which is driven by the crankshaft and, for each revolution, generates pulses which can be sensed by the sensor, and deviations of the rotational speed from a predefined setpoint value are compensated for by an actuator which acts on the fuel injection of the engine.
Methods of this type are primarily used in diesel engines which, e.g., in comparison with spark-ignition engines, for example, inevitably require the idling speed to be regulated and the maximum rotational speed to be limited. The injection pumps used for this inject more or less fuel into the cylinders as a function of the existing rotational speed, for which purpose they are adjusted via appropriate control devices. In large diesel engines, e.g., in the case of large two-stroke engines which can often be found on board ships, each cylinder is provided with its own individual injection pump, which receives a hydraulic power boost because of the necessary, high actuating forces. The large two-stroke engines used on board ships because of their proven reliability, are generally designed as in-line engines having four to twelve cylinders, and operate in a nominal rotational speed range of 50 to 120 rev/min. Since marine diesel engines are reversible, that is to say, can run in both directions of rotation, the camshafts have cams for forward and reverse running, which operate the individual injection pumps.
Due to the low nominal rotational speeds, the result is minimum rotational speeds of 12 to 25 rev/min, which are characterized by rough running. The cause of this is rotational speed fluctuations, which are brought about by periodic interferences such as the individual ignition operations, and by temporary irregularities such as load surges, changes in the drive torque, ignition misfires or other malfunctions of the injection pumps. In order to avoid the engine stopping or overrunning, the rotational speed fluctuations are compensated for by readjusting to a constant rotational speed.
Methods for maintaining predefined setpoint rotational speed values are known which correct both static deviations, that is to say periodic rotational speed deviations at constant load, and dynamic deviations, that is to say, temporary setpoint value deviations, such as during start-up between idling and full load. Thus, European Patent No. 0 481 983 B1 describes a method of regulating the rotational speed of a slow-running multicylinder diesel engine in which, for each of the cylinders, angular positions of the crankshaft are defined which represent the starting angle and finishing angle of an angular range located before the top dead center of the cylinder. For these angular ranges, which are detected by means of a sensor for corresponding pulse-generating marks which rotate with the crankshaft, an actual value is measured continuously, which indicates the average speed at which the crankshaft passes through this angular range. This actual value,.is fed to a xe2x80x9crapidxe2x80x9d, that is to say, proportionally acting, controller, which acts on the charging level (volumetric efficiency) of the cylinder assigned to the corresponding angular range. In addition, a speed of the crankshaft averaged over a number of these angular ranges is measured and fed to a xe2x80x9csluggishxe2x80x9d, that is to say, an integrally or proportionally-integrally acting controller, which is used to preset the charging level of all the cylinders. In order to compensate for occurring disturbances as early as possible through correction of the charging level, the position of the angular ranges determined by the starting and finishing angles is also adjusted as a function of the rotational speed of the crankshaft.

It proves to be a drawback of this method, that the regulation of low rotational speeds of below approximately 25 minxe2x88x921 cannot be carried out satisfactorily. In addition, the dynamic rotational speed deviations occurring as a result of load surges cannot be compensated for better than with conventional rotational speed controllers. Furthermore, rough movements of the charging linkage can be observed under constant load if the injection pumps are not adjusted exactly. In the event of malfunctioning of individual injection pumps, large periodic movements of the charging linkage can be seen, which result from the fact that the xe2x80x9cquickxe2x80x9d controller is continuously attempting to correct the dip in the rotational speed of the respective cylinder caused by this.
The present invention is based on the object of developing a method of regulating the rotational speed of multicylinder internal combustion engines to the effect that, whilst avoiding the above-described disadvantages, precise and rapid detection of, the rotational speed may be achieved in a simple and cost-effective manner, permitting stable regulation to a constant rotational speed in the event of periodic and temporary rotational speed fluctuations.
This object of the present invention is achieved by continuously measuring the time required for a fixed sequence of successively sensed pulses and using it to form equidistant, uncorrected actual rotational speed values, the number of successively sensed pulses being defined as a function of the number of cylinders in the engine and the number of pulses that can be generated by the pole wheel per revolution; by measuring the time required for the number of sensed pulses corresponding to one complete revolution of the pole wheel and using it to form an average rotational speed; by smoothing the difference between the average rotational speed and each uncorrected actual rotational speed value in order to form corrected actual rotational speed values, by limiting its magnitude and adding it to the uncorrected actual rotational speed value; and by comparing the corrected actual rotational speed values with a predefined setpoint rotational speed value and feeding them to a controller, whose output variable is used to control the actuator.
Using such a method, both temporary and periodically occurring rotational speed fluctuations are identified rapidly and precisely, and compensated for by the controller through appropriate control of the actuator acting on the fuel injection.
It is of particular advantage to define the number of successively sensed pulses as the rounded quotient of the number of pulses that can be generated by the pole wheel per revolution, and the number of cylinders in the engine. This offers the advantage that a sequence of sensed pulses can be assigned to a cylinder, with the consequence that the actual rotational speed value formed corresponds to the averaged rotational speed of the crankshaft over a rotational angular range, in which two cylinders firing one after another assume the same position, e.g. top dead center. In this manner, the actual determined rotational speed value contains no influences occurring periodically during one working cycle, such as the ignition processes.
According to an advantageous further refinement of the present, the uncorrected actual rotational speed values are formed at intervals of two successive pulses. This means that, in the event of a number of more than two pulses of a set pulse sequence, the uncorrected actual rotational speed values formed are assigned to overlapping angular ranges of the crankshaft, so that the rotational speed can be detected precisely and rapidly because of the resulting high number of actual rotational speed values.
According to an additional feature of the present invention, each sequence of sensed pulses is assigned a measured value store, in which the time intervals of the successively sensed pulses are summed to form an uncorrected actual rotational speed value. This permits each sensed pulse to be assigned simultaneously, and in a simple manner, to a number of pulse sequences resulting from the corresponding actual rotational speed values, on account of the overlapping angular ranges of the crankshaft. This can be implemented in a conventional manner using a process computer having a nonvolatile memory. Expediently, following the formation of an uncorrected actual rotational speed value, the content of the measured value store is kept, at least for one revolution, before being deleted and assigned to a new sequence of sensed pulses. As a result, the measured value can be used for calculating the averaged rotational speed value. In addition, further evaluations of the rotational speed variation (characteristic curve), such as with regard to detecting misfiring or acceleration occurrences, are then possible.
The pulses are advantageously sensed using two sensors arranged to be offset by 90xc2x0, in order to detect the direction of rotation of the crankshaft, which can be reversed particularly in marine diesel engines, and as a result, to establish the sign when counting the pulses. It is also advantageous to use as the pole wheel, a gearwheel having teeth arranged in a uniformly distributed manner along its circumference, so that the sensing of pulses is ensured in a simple manner. According to a further advantageous feature of the present invention, a pulse is generated in the direction of rotation by both the leading and trailing flanks of a rotating gearwheel, in order to obtain a high number of pulses which can be sensed. A particularly advantageous implementation of the method to this end results in the case of approximately symmetrical tooth flanks, so that constant time intervals between the individual measurements can be achieved. The measurement accuracy is not influenced by the arrangement of the teeth or sensors.
Advantageously, the pulses are sensed equidistantly, so that digital sampling control can be used and a mark-to-space ratio of approximately 1 results. For this purpose, it is also advantageously proposed to use a gearwheel having a ratio of tooth thickness to gap width on the pitch circle of approximately 1.
According to a feature of the present invention, the pulses are detected by means of inductive, capacitive or optical position sensors in order to achieve a simple and cost-effective measurement set-up. Alternatively, use can also be made of incremental position sensors driven without slip or no other digital or analog detector circuits coupled to the crankshaft.
In order to regulate to constant rotational speeds in a stable manner, a development of the invention uses a proportional-integral controller with a derivative action time, which corrects the occurring rotational speed deviations. Finally, it is proposed to use adaptive control in order, for example, to take into account influences arising from the environment.